The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external source code and user-provided issue descriptions, which creates a surface for indirect prompt injection. A malicious codebase being analyzed could contain comments or patterns designed to override the agent's instructions.
Ingestion points: Target source files and issue descriptions passed to the /sc:analyze command.
Boundary markers: Absent. The skill instructions do not specify delimiters or provide warnings to the agent to treat analyzed content strictly as data.
Capability inventory: The skill utilizes Bash for command execution, Write for file modification, and mcp__rube__RUBE_MULTI_EXECUTE_TOOL for sending external notifications (e.g., Slack, Jira).
Sanitization: Absent. There are no instructions for sanitizing or filtering instructions that might be embedded within the files being scanned.
[COMMAND_EXECUTION]: The skill explicitly includes Bash in its tool coordination list for external tool execution and diagnosis. While standard for an analysis tool, this capability represents a significant impact vector if the agent is successfully manipulated via indirect prompt injection.

sc-analyze